perf-moe-hardware-configs

Original：🇺🇸 English

Translated

Representative MoE training playbooks by hardware platform and model family. Summarizes rounded throughput bands, parallelism patterns, and common tuning stacks.

2installs

Sourcenvidia/skills

Added on2026-05-22

NPX Install

npx skill4agent add nvidia/skills perf-moe-hardware-configs

SKILL.md Content

View Translation Comparison →

MoE Hardware Configuration Reference

Stable docs: @docs/training/moe-optimization.md Card: @skills/perf-moe-hardware-configs/card.yaml

Quick Platform Playbook

Platform	Typical MoE strategy	What usually matters most
H100	DeepEP + stronger PP + moderate TP	communication overlap and PP efficiency
B200	DeepEP + MXFP8 + careful PP layout	container quality and tuned comm settings
GB200	HybridEP + partial CUDA graphs + CPU cleanup	host overhead, topology-aware dispatch, memory headroom
GB300	HybridEP + newer FP8 and kernel stack	same GB200 playbook, usually with a higher ceiling

Rounded Performance Bands

These are intentionally rounded so the document stays durable as the tracker moves. Treat them as planning ranges, not exact promises.

Workload family	Hardware	Typical band	Representative shape
DSV3, large-scale	H100	low-to-mid hundreds TFLOPS/GPU, high-teens MFU	TP2, EP64, PP8, DeepEP
DSV3, large-scale	B200	high-hundreds TFLOPS/GPU, mid-teens MFU	TP1, EP32, PP8, DeepEP
DSV3, large-scale	GB200	around 1K TFLOPS/GPU, low-20s MFU	TP1, EP64, PP4, HybridEP
DSV3, large-scale	GB300	above the GB200 band, often mid-20s MFU	TP1, EP64, PP4, HybridEP
Qwen3 235B	H100	low-300s TFLOPS/GPU, around 30% MFU	TP2, EP32, PP8, DeepEP
Qwen3 235B	GB200	high-hundreds TFLOPS/GPU in tuned runs	TP1 or TP2, EP32-64, PP4, HybridEP
Qwen3 30B	H100	low-200s TFLOPS/GPU	TP1, EP8, PP1, DeepEP
Qwen3-Next 80B	GB200	low-300s TFLOPS/GPU in BF16-class runs	TP1, EP32, PP2, HybridEP

Representative Config Families

DSV3 on H100

text

Dispatcher: DeepEP
TP=2  EP=64  PP=8  VPP=4
Routing: force balance
Recompute: light-to-moderate selective recompute
Priority: overlap communication and keep PP efficient

DSV3 on B200

text

Dispatcher: DeepEP
TP=1  EP=32  PP=8  VPP=2 or similar
Precision: MXFP8-class
Recompute: selective recompute around MLA up-projection and MLP-side modules
Priority: container quality, PP layout, and DeepEP SMS tuning

DSV3 on GB200 or GB300

text

Dispatcher: HybridEP
TP=1  EP=64  PP=4  VPP=4
Precision: MXFP8-class
CUDA Graph: attn + moe_router + moe_preprocess
Priority: HybridEP, CPU optimization, and graph-friendly static shapes

Qwen3 235B on H100

text

Dispatcher: DeepEP
TP=2  EP=32  PP=8  VPP=4
Recompute: norm and activation-side selective recompute
Priority: communication overlap and router-path cleanup

Qwen3 235B on GB200

text

Dispatcher: HybridEP
TP=1 or 2  EP=32 to 64  PP=4
CUDA Graph: attn + moe_router + moe_preprocess
Recompute: moe_act, mlp, or norm depending on memory pressure
Priority: balance throughput against memory headroom

Qwen3-Next 80B on GB200

text

Dispatcher: HybridEP
TP=1  EP=32  PP=2  VPP around 4
CUDA Graph: attn + moe_router + moe_preprocess
Priority: pipeline layout and grouped GEMM quality

Cross-Cutting Patterns

PP layout

```
E
```
= embedding
```
t
```
= transformer
```
m
```
= MTP
```
L
```
= loss
```
|
```
= stage boundary

The biggest platform difference is usually not just the dispatcher. It is the combination of dispatcher, PP shape, and whether VPP keeps each stage balanced.

Recompute strategy

Memory pressure	Starting point
low	none or a very narrow selective set
moderate	`moe_act` , `mlp` , `norm` , or similar selective modules
high	model-specific up-projection plus selective MoE and MLP modules
extreme or long-context	full recompute only if the selective path still does not fit

Environment variables

bash

CUDA_DEVICE_MAX_CONNECTIONS=1
CUDA_DEVICE_MAX_CONNECTIONS=32   # common when EP overlap and CUDA graphs are combined
PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True
NCCL_GRAPH_REGISTER=0

CPU-side tuning

On GB200 and GB300, CPU affinity and general host-overhead cleanup can move the needle almost as much as a dispatcher swap. Treat them as first-class tuning work, not as afterthoughts.

Pitfalls

Do not cargo-cult a tracker row: the winning config usually depends on routing mode, container, and PP layout as much as on hardware name.
Container quality matters: large regressions can come from the software stack rather than the model recipe.
VPP must be intentional: a bad VPP split can erase the gain from a better dispatcher.
Compare absolute throughput, not only MFU: MFU can mislead when switching between BF16, FP8, and other precision modes.
Force-balance routing is the safer benchmark default: keep routing mode fixed when comparing hardware or dispatcher stacks.

perf-moe-hardware-configs

NPX Install

Tags

SKILL.md Content

MoE Hardware Configuration Reference

Quick Platform Playbook

Rounded Performance Bands

Representative Config Families

DSV3 on H100

DSV3 on B200

DSV3 on GB200 or GB300

Qwen3 235B on H100

Qwen3 235B on GB200

Qwen3-Next 80B on GB200

Cross-Cutting Patterns

PP layout

Recompute strategy

Environment variables

CPU-side tuning

Pitfalls